The new AgentCore runtime: Elastic, optimized, and consistently fast starts
Today we are announcing the new AgentCore runtime, a capability of Amazon Bedrock AgentCore built for the speed, flexibility, and cost efficiency that production agents demand. It reclaims memory as sessions release it and delivers consistent cold starts regardless of image size or concurrency.
Amazon Bedrock AgentCore has introduced a new runtime designed to improve the speed, flexibility, and cost-efficiency for production agents. This new runtime addresses two main challenges faced by agents: memory usage and startup time consistency.
The new runtime optimizes memory management by reclaiming unused memory immediately when a session ends, rather than holding it at peak usage. This change eliminates the need for agents to pay for peak memory usage throughout their entire run. Additionally, the runtime starts each session from a small, efficient memory profile instead of a full provisioned footprint. Additional memory is allocated and paged in on demand as needed, allowing the platform to scale down to zero when an agent is idle, further reducing costs.
Startup times remain consistent regardless of container size or agent concurrency. Sessions now start faster, even for larger or more concurrent agents, by avoiding the lengthy cold start process that involves booting a fresh environment, pulling the image, and initializing the agent. This improvement ensures a better user experience, particularly for agents that users interact with frequently.
The new AgentCore runtime effectively resolves the issues of memory wastage and inconsistent startup times, making it easier for developers to build, connect, and optimize agents securely at scale.
Written by urgent.news from AWS Machine Learning's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.
